Socioeconomic and Healthcare Predictors of Life Expectancy Across Countries: Evidence from 2019, 2021, and 2022

Abstract

Life expectancy reflects the health status of a population and the level of social development in a given country. This study utilizes data from 217 countries and regions for the years 2019, 2021, and 2022 to analyze the relationships between life expectancy and various factors, including income, education, medical resources, sanitary conditions, healthcare expenditure, fertility rates, and urbanization. The year 2019 serves as a pre-COVID-19 baseline, while 2021 and 2022 allow for the observation of any shifts in these relationships during the pandemic. The study employs linear regression and random forest models for prediction, with results evaluated through cross-validation and out-of-sample residual analysis. The findings indicate that basic sanitary facilities, GDP per capita, and fertility rates are strong predictors of life expectancy across all three years. Countries with higher coverage of sanitary facilities and higher income levels generally exhibit higher life expectancy, whereas those with higher fertility rates tend to have lower life expectancy. Other factors—such as the number of physicians, female secondary school enrollment rates, and the proportion of the urban population—show weaker predictive power or lack stability across the years examined, while healthcare expenditure is a moderate predictor. While both models largely agree on the primary determinants, the tuned random forest has somewhat better out-of-sample accuracy in all three years. In some countries, actual life expectancy is well below the model-predicted values, suggesting that income and medical resources alone cannot fully account for cross-national disparities in health. These results suggest that, in efforts to increase life expectancy, governments should prioritize basic sanitary facilities and public health services, while also improving education, maternal and child health, and the efficiency with which medical resources are utilized. Policymakers should also pay close attention to countries where actual life expectancy consistently falls below predicted levels and further investigate factors such as the distribution of medical services, the burden of infectious diseases, social inequality, and public governance. It should be noted that this study identifies predictive associations and does not establish a causal relationship between these factors and life expectancy.